[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <877drxp3i5.fsf@kmk-computers.de>
Date: Sun, 11 Oct 2020 10:32:02 +0200
From: Kurt Kanzenbach <kurt@...-computers.de>
To: Florian Fainelli <f.fainelli@...il.com>,
Andrew Lunn <andrew@...n.ch>,
Vivien Didelot <vivien.didelot@...il.com>,
Vladimir Oltean <olteanv@...il.com>
Cc: "David S. Miller" <davem@...emloft.net>,
Jakub Kicinski <kuba@...nel.org>,
Rob Herring <robh+dt@...nel.org>, netdev@...r.kernel.org,
devicetree@...r.kernel.org, kurt@...utronix.de
Subject: Re: [PATCH net-next 1/2] dt-bindings: net: dsa: b53: Add YAML bindings
On Sat Oct 10 2020, Florian Fainelli wrote:
> On 10/10/2020 9:46 AM, Kurt Kanzenbach wrote:
>> Convert the b53 DSA device tree bindings to YAML in order to allow
>> for automatic checking and such.
>>
>> Suggested-by: Florian Fainelli <f.fainelli@...il.com>
>> Signed-off-by: Kurt Kanzenbach <kurt@...-computers.de>
>
> Thanks for making this change, there are quite a few warnings that are
> going to show up because the binding was defined in a way that it would
> define chip compatible strings, which not all DTS files are using.
Oh, I didn't know there is a second make command for doing the actual
check against the dtbs. I've just used `make dt_binding_check'.
So, it seems like a lot of the errors are caused by the include files
such as
[linux]/arch/arm/boot/dts/bcm5301x.dtsi
srab: srab@...07000 {
compatible = "brcm,bcm5301x-srab";
reg = <0x18007000 0x1000>;
status = "disabled";
/* ports are defined in board DTS */
};
The nodename should be "switch" not "srab" as enforced by
dsa.yaml. Furthermore, some DTS files are not adding the chip specific
compatible strings and the ports leading to more errors.
There are also some minor errors regarding the reg-names and such for
specific instances.
How should we proceed? Adding the missing compatible strings and ports
to the DTS files? Or adjusting the include files?
> I don't know if Rob would be comfortable with taking this until we
> resolve all warnings first.
Probably not. We should fix the existing device trees first.
Thanks,
Kurt
Download attachment "signature.asc" of type "application/pgp-signature" (866 bytes)
Powered by blists - more mailing lists